Because of the increase in data intensive applications, the demand for bandwidth in communications has been growing tremendously. In response, the installed capacity of telecommunication systems has been increasing by an order of magnitude every three to four years since the mid 1970s. Much of this capacity increase has been supplied by optical fibers that provide a four-order-of-magnitude bandwidth enhancement over twisted-pair copper wires.
To exploit further the bandwidth of optical fibers, two key technologies have been developed and used in the telecommunication industry: optical amplifiers and wavelength-division multiplexing (WDM). Optical amplifiers boost the signal strength and compensate for inherent fiber loss and other splitting and insertion losses. WDM enables different wavelengths of light to carry different signals parallel over the same optical fiber. Although WDM is critical in that it allows utilization of a major fraction of the fiber bandwidth, it would not be cost-effective without optical amplifiers. In particular, a broadband optical amplifier that permits simultaneous amplification of many WDM channels is a key enabler for utilizing the full fiber bandwidth.
With the advent of erbium-doped fiber amplifiers (EDFAs) around 1990 to replace electronic repeaters, the capacity of telecommunication systems has since been increased by almost two orders of magnitude. Although EDFAs have had a significant impact in the past five years, they are not without problems. As shown in FIG. 1a, there are two main low-loss telecommunications windows in silica-based optical fibers at wavelengths of 1.3 μm and 1.55.μm. EDFAs work only in the 1.55 μm window. Yet, most of the terrestrial fibers installed in the United States during the 1970s and up through the mid 1980s are designed for operation at 1.3 μm, and thousands of miles of 1.3 μm terrestrial fibers have already been laid. This presents major difficulties in upgrading to the higher bandwidth EDFA technology. In the prior art, some have sought to combine EDFAs with dispersion compensators in an effort to correct the wavelength mismatch. However this approach does not permit further upgrading based on wavelength-division-multiplexing, and therefore is not seen as the best solution. Others are experimenting with new glass formulations that might provide the advantages of EDFAs at the shorter 1.3 μm wavelength. However, currently no glass formulation has proven to be commercially viable.
Aside from the wavelength mismatch, EDFAs are also inherently prone to signal loss when the pump laser fails. EDFA is a system of the type known as a “three-level” system that does not allow the optical signal to pass through unless its pump laser is operative. Reliance on the “three-level” system could have catastrophic consequences for the reliability of fiber networks.
Stimulated Raman scattering amplifiers are advantageous over EDFAs because they can operate in both optical communication windows and, in fact, over the entire transparency window of optical fibers. Moreover, the stimulated Raman scattering amplifier is a “four-level” system that simply provides no gain when its pump laser is off, but otherwise allows the optical signal to pass through the system. Stimulated Raman scattering amplifiers are based on nonlinear polarization of the dielectric silica host, and are capable of cascading to higher Raman orders or longer wavelengths. However, there is a significant problem with Raman amplifiers that has not heretofore been really overcome. Virtually every light source or pump produces some intensity fluctuation. When Raman amplifiers are allowed to cascade through several orders, the pump source intensity fluctuations are combinatorially multiplied, and very rapidly result in enormous intensity fluctuations that have made systems virtually unusable. Compounding this problem, the gain produced by this nonlinear response is proportional to instantaneous pump intensity. Thus there is no opportunity to “average out” intensity fluctuations over time. Moreover, the gain produced by Raman scattering is, itself, an exponential effect. All of these properties have lead most to conclude that stimulated Raman scattering amplifiers and cascade lasers are not suitable in general-purpose telecommunication applications.
Aside from the fluctuation problems above, several other issues also need to be addressed in order to achieve usable broadband stimulated Raman scattering amplifiers. In the prior art, a cladding-pumped fiber laser has been used as a pump source for Raman amplifiers. A commercial unit delivers 9 W of single-transverse-mode output at 1100 nm with a spectral width of 4 nm. The fiber used in this laser is a rare-earth-doped, double-clad fiber. As depicted in FIG. 14a, cavity mirrors are applied to the fiber ends. The mirror applied to the input end is highly reflective at the lasing wavelength of 1100 nm, while a low-reflectivity mirror or grating is applied to the output end of the fiber. The gain band for ytterbium doped fiber is roughly between 1030 nm and 1160 nm, but using a grating at the fiber end to select one particular wavelength yields a bandwidth of about 4 nm.
While this cladding-pumped fiber laser is already quite broad in bandwidth because of multiple longitudinal modes in the cavity, it would be desirous to further broaden the pump wavelength range to achieve broadband Raman gain. The broader pump bandwidth is also advantageous to avoid reflections associated with stimulated Brillouin scattering in the gain fiber of the Raman amplifier.
Polarization controllers (PCs) are used in almost all Raman amplifiers to regulate polarization states. A fiber based PC is typically constructed using quarter-wave loops of optical single-mode fiber mounted in such a way as to allow precise rotation of the loops about a common tangential axis. Each loop is designed to function as a quarter-wave retarder for the wavelength range of interest. By rotating a loop about its tangential axis, the loop's birefringence is rotated. Combining three or four loops in series increases the wavelength range and adjustment range of the controller and enables complete and continuous polarization adjustability. However, as the temperature changes, the fiber birefringence changes and the mechanical setting of the PCs may also be perturbed. As a result, the PCs may ruin the “turn-key” operation of the amplifier because they could require periodic readjustment with changing environmental conditions.
In the past attempts at applying Raman amplifiers to analog signal amplification, it was discovered that a major limitation arises from the noise associated with Double Rayleigh Scattering (DRS). Stimulated Rayleigh scattering refers to light scattering due to induced density variations of a material system. More specifically, stimulated Rayleigh refers to the scattering of light from isobaric density fluctuations.
Stimulated Rayleigh scattering gives rise to a backward traveling wave that is at the same center frequency as the signal input, somewhat broadened by the Rayleigh linewidth (defined as a reciprocal to characteristic decay time of the isobaric density disturbances that give rise to Rayleigh scattering). For example, J. L. Gimlett, et al., IEEE Photonics Technology Letters, Vol. 2, p.211 (March 1990) disclosed that the Rayleigh scattering can be modeled as a Rayleigh mirror with a prescribed reflectivity. DRS refers to a second stimulated Rayleigh scattering event that scatters the backward traveling wave back into the original signal, thereby leading to interference with the original signal, cross-talk, and increased uncertainty of the amplitude (i.e., noise). Also, the DRS is proportional to the pump intensity, the signal intensity, and the length of the gain fiber. Therefore, the DRS noise source is a direct consequence of requiring high pump powers and long interaction lengths due to the inefficiency of the Raman amplification process.
Prior art has shown that insertion of an optical isolator midway through the amplifier and the use of two WDMs to guide the pump radiation around the isolator can reduce the DRS effect. In effect, the amplifier is split into two parts and the net gain is accumulated through both sections, but the isolator reduces the DRS in half. Although this technique has been used for high gain EDFAs and in ring designs of Raman amplifiers, it increases the complexity and cost of the amplifier considerably due to the need for two additional WDMs and one isolator.
As shown from the attenuation curve for fibers in FIG. 1a, there are two low-loss windows for telecommunications. In the prior art, EDFA technology has been developed to make full use of the 1.5 μm window. Since Raman amplification can be obtained over the entire transparency range for optical fibers, Raman amplification can be applied to both the 1.3 μm and 1.5 μm windows. Because future communication applications will demand the broadest bandwidth available over the existing fiber base, to fully utilize optical fiber's bandwidth, it is desirable to have an amplifier which will use both telecommunications windows and operate with WDM simultaneously.